It's all about proper application, as stated by those that know. If a part is more cheaply or easily made by machining a readily available shape, then, of course, you'd machine it. The naysayers only look to find what the new technology has for a weakness and then point out that weakness where another method is stronger and attempt to use it as a fallicious proof that the new technology is completely deficient. Every method has it's strengths and weaknesses. New methods will be used where their strengths excell. As the technology improves or changes or becomes hybridized, it will phase out the the older methods where they are more difficult. However, it will not eliminate any methods. Some methods, although more cheap in manufacturing, are too expensive for on-off projects.
It may be cheaper for someone to drill and file out a hole in a lensboard by himself than to drive to a machine shop and hire a machinist with a lathe to do a quick bore. You make due with what you have. If it works, then it works--you can't argue with something that works. The dude printed a camera and it works--say whatever you want, it's a success and you can't argue that.
The 3d printing will allow the people without the equipment and skills to produce items that they would otherwise not produce. Items can be perfected in this way iteratively--the way REAL engineers design things.